JOURNAL OF RESEARCH IN SCIENCE TEACHING VOL. 29, NO. 1, PP.

17-34 (1992)

Using Examples and Analogies to Remediate Misconceptions in Physics:

Factors Influencing .Conceptual Change

David E. Brown
Department of Curriculum and Instruction, College of Education, University of
Illinois at Urbana-Champaign

Abstract
Examples are often recognized as important in teaching conceptual material, yet little
research has been done concerning the best use of examples in attempts to remediate misconceptions.
This study questions the effectiveness of a traditional teaching-by-example technique. Results
indicate that when students hold a misconception, presenting a principle with suppodng examples
to show the range of application of the principle may be ineffective. Rather, it appears that
examples are more effective when they help students draw on and analogically extend existing
valid physical intuitions in constructing a new conceptual model of a target situation. To help
students in this constructive effort, first, the examples used must be understandable and believable
to the students, not simply to the teacher or textbook author. Second, even when an example
is compelling to the student, it may not be seen as analogous to target problems drawing out a
misconception. In that case, analogy relations may need to be explicitly developed. Third,
qualitative, visualizable models may need to be developed which give mechanistic explanations
for phenomena.

Recent research indicates that students come to the science classroom with a
number of naive conceptions which can inhibit the learning and understanding of
certain concepts (Driver & Easley, 1978; Driver & Erickson, 1983; McCloskey, 1983;
Mckrmott, 1984). Many of these misconceptions’ are widespread and have a detrimental
effect on problem solving (Champagne, Gunstone, & Klopfer, 1982), course performance

’ In this article I use the term “misconception” to refer to students’ ideas which are incompatible with
currently accepted scientific knowledge. To be sure, these. conceptions should be respected as creative
constructions of the individual which in many cases are adaptive and successful for dealing with the practical
world. These naive conceptions do, however, present significant barriers to learning a subject like Newtonian
mechanics and as such need to be addressed as difficultiesfrom the perspective of the content domain being
taught.

8 1992 by the National Association for Research in Science Teaching

Published by John Wiley & Sons, Inc. CCC 0022-4308/92/001017- 18W.00
18 BROWN

(Champagne, Klopfer, & Anderson, 1980; Halloun & Hestenes, 1985), and, by definition,
conceptual understanding of the material. These naive conceptions also appear to be
resistant to change via traditional instructional approaches (Sjoberg & Lie, 198 1;
Viennot, 1979; Clement, 1982; Halloun & Hestenes, 1985; Brown & Clement, 1987).
One potential resource for the task of remediating misconceptions is research
which has been conducted on concept learning. A recurrent theme in this literature is
the appropriate use of concrete examples (e.g., Tennyson & Park, 1980; Mervis 8:
Pani, 1980). A recent cognitive view of how multiple examples or analogies2enhance
learning is that the examples provide the input to an inductive process, leading to an
abstract schema which contains only those features crucial to the concept. This abstract
schema can then be applied to other cases which match the essential features of the
schema (Gick & Holyoak, 1983; Holland, Holyoak, Nisbett, & Thagard, 1986; Gentner,
1983, 1989). However, when students hold a misconception, competing schemas may
interfere with the process of schema induction (Kaiser, Jonides, & Alexander, 1986).
Because of such prior conceptions, it may be necessary to show students how
certain examples are like prototypical cases (Rosch, 1973; Rosch & Mervis, 1975)
rather than simply telling them that certain situations are examples of a concept. This
latter method of telling students which examples illustrate the application of a concept
may be appropriate when the students have no initial conceptions, but it may be
inappropriate and ineffective when the students already have conceptions about the
target domain.
This raises the question of how examples should be used for domains in which
students hold misconceptions. Clement (1989) presents cases of noninductive learning
by physicists and argues that the process of scientific discovery is better described as
a process of rational model construction rather than purely as inductive generalization.
In particular, he observed the physicists using multiple analogies in this process of
model construction and evaluation. For example, the physicists appeared to use in-
termediate or “bridging” analogies to evaluate the soundness of analogical models.
In this study I pose two questions. (a) Can students be productively engaged in
a similar process of model construction and evaluation by means of multiple analogies?
(b) Would such an approach be more effective than an inductive approach in which
multiple examples illustrate the application of a concept or principle? An interview
study was conducted in order to explore student response to these two different uses
of examples and to provide indications of student reasoning while they interact with
the example-based explanations.

Method

Subjects
Twenty-one high school chemistry students volunteered to be interviewed for
approximately 45 minutes each. The students had not :yet taken physics but were

* For the purposes of this article, “example” and “analogy” are used to refer to physical situations
which are in some way related to each other. As used here, when the relationship is primarily that each of
the situations can be assimilated under the same abstract principle, “example” is the preferred term; when
they are primarily related in that they have structural and/or functional similarities, “analogy” is the prefemed
term.
 USING EXAMPLES To REMEDIATE MISCONCEPTIONS 19

representative of students who might take physics the following year. The students’
chemistry teachers were asked to categorize them as having either a relatively easy
or a difficult time with conceptual material. Combined with the information about
their chemistry course level (advanced or standard), each student was assigned to
one of four subgroups. Half of the students in each subgroup were chosen at random
to receive one of the following explanations, and the other half received the other
explanation.

The Teaching Instrument

Explanations. Both example-based explanations (a text excerpt and a bridging

explanation) consisted of seven short paragraphs. After reading each paragraph, students
were asked questions to focus them on the content of the paragraphs and encourage
active involvement. (See Appendix A for the complete text of both explanations.)
Text Excerpt. The text excerpt contains a verbatim excerpt from a popular and
innovative high school physics textbook (Rutherford, Holton, & Watson, 1981). The
excerpt presents a number of examples of Newton’s third law, the law of action and
reaction. Some of the examples used are a finger pressing on a stone (one of Newton’s
own examples; the stone presses back on the finger), an athlete running (the ground
pushing forward on the athlete is responsible for her motion), and a rifle kicking.
Added to this verbatim excerpt were some diagrams illustrating the examples, two
sentences at the beginning relating the explanation to the situation of a book resting
on a table, and a final paragraph explicitly stating that Newton’s third law indicates
that the table is exerting an upward force on the book.
Bridging Explanation. The bridging explanation also made use of concrete situations
from the students’ experience. Unlike the text excerpt, the examples formed a connected
sequence, starting from an “anchor” (a situation in which most students believe there
is an upward force, in this case a hand pressing down on a spring), through intermediate
situations (e.g., a flexible board between two sawhorses), to the target situation of a
book on a table. Thus this explanation shows, by means of a connected sequence of
examples, where the force comes from-the microscopic compression or bending of
the table.
This explanation is designed to (a) ground understanding on an anchoring intuition
that the student already possesses; (b) help the student develop a conviction that the
target problem is in fact analogous to the anchoring case; and (c) help the student
build a qualitative, mechanistic model of rigid objects (compressing and/or bending
on a microscopic scale) that is also based on the anchoring intuition. In contrast, the
text examples seem intended to show part of the range of application of Newton’s
third law. The differences between the bridging and text excerpts are illustrated in
Figures 1 and 2.
Pre- and Postquestions. Each student received a set of three prequestions and
five postquestions (three identical to the prequestions plus two additional questions--
see Appendix B). Each of the problems (except the table problem) had two parts. The
first part asked about the existence of a force from a static object, and, if the student
answered that a force exists, the second part asked the student to compare the magnitude
of this force with that of another force. The table problem asked only about the existence
of a force from the table.
 Induction from examples

I I T W

Examples Target Problem

Figure 1. Text excerpt. Induction from multiple examples to generate or support an abstract schema which can then be applied
to new cases.
 Anchoring Situation Target Problem
Figure 2. Bridging explanation. Comparison of analogies bridging between an anchoring situation and a target problem, leading
to model construction.
22 BROWN

On each question the students were asked to rate (a) their confidence in the answer
given, and (b) how much sense the answer choices made. Having confidence in an
answer and having an answer make sense were carefully distinguished for the student.
The main reason for this distinction is to try to uncover what students intuitively believe
is correct rather than what they may conjidently believe is correct because they happen
to remember something in a rote fashion.

Quantitative Results
Of the 21 students, 14 initially maintained that a table does not exert an upward
force on a book resting on it. Of these, 7 received the text excerpt, and 7 received
the bridging explanation. Table 1 shows the number of' correct responses for each
group on the pre- and post-questions. (Since Questions 4 and 5 were asked only after
the explanation, they do not have pre-explanation scores, and Table 1 shows only the
post-scores.) Correct answers on each part of the two-pat questions (Questions 2-5)
were scored as counting one point, and comparisons were made for the scores on each
question. The Mann-Whitney U test was used for the pairwise comparisons.
All of the bridging students who initially answered the table problem incorrectly
answered the post-question about the book on the table comctly and with hgh confidence
(the average confidence score was 2.79 out of 3). They also indicated that this answer
made a great deal of sense to them (the average sense rating was 4.64 out of 5), and
their performance on other post-questions was quite encouraging. Particularly encouraging
is the fact that six of the seven students answered both pads of the steel blocks problem
correctly, a difficult transfer problem which draws out Ihe strong intuition in many
students that force is a property of objects. Many students who are given this problem

Table 1
Number of Correct Responses for Each Problem
(Students Initially Answering the Table Problem
Incorrectly)
-
Text excerpt Bridging
(n = 7) i:n = 7)
Questions Pre Post Re Post

Tablea 0 2 0 7*
Goatb 3 6 5 13
Mosquitob 1 4 2 14*
Two boxesb - 5 - 13**
Steel blocksb - 3 - 13++

"Maximum = 7 (one part question).

bMaximum = 14 (two part question).
* p < .01 (comparing pre/post differences).
** p < .05 (comparing post-scores).
tt p < .OI(comparing post-scores).
Note. Probability levels obtained from the Mann-Whitney U
test.
 USING EXAMPLES TO REMEDIATE MISCONCEPTIONS 23

answer that block A (a 200-pound steel block resting on a 40-pound steel block) exerts
the larger force since it is heavier. On a recent high school diagnostic test, after a full
year of traditional instruction in physics, only 40 of 78 students answered this problem
correctly (Brown, 1989).
By contrast, of the seven text students who initially answered the table problem
incorrectly, five answered the table problem incorrectly after reading the explanation,
even though the explanation had explicitly stated the correct answer to this problem.
They performed similarly on other post-questions. In particular, none of them answered
both parts of the steel blocks problem correctly.

Descriptive Observations and Discussion

The bridging explanation and the text excerpt differ on a number of dimensions,
making comparison between the two methods difficult. In particular, it is not possible
to isolate factors responsible for any differences with a purely quantitative analysis.
However, qualitative analysis of the interview protocols enables at least preliminary
isolation of some factors. Several possible reasons are explored in this section for the
observed differences in student response to the two explanations.
The students were videotaped for in-depth study of the interactions with the
explanations and of the reasons students gave for their answers. In this section, further
evidence of the superiority of the bridging students’ understanding is provided, and
some factors which may have contributed to the differences between the groups are
proposed.

Comparative Effectiveness of the Explanations

Table Post-question
Text Excerpr Students. The text excerpt presented a principle which was supported
by a number of examples from the students’ experience and which explicitly stated
that the book on the table was another example of the stated principle. Despite this,
the majority of the text students continued to maintain the absence of a force from the
table. There are at least two possible reasons for this: (a) The students did not realize
that the principle explicated in the text excerpt (Newton’s third law) should apply to
the table situation; or (b) they realized the principle should apply, but they simply
refused to accept this conclusion. Students’ statements provide support for the second
reason.
All of the text students explicitly stated, in their reasons for their answers, that
the text excerpt indicated that the table exerts an upward force. Thus, all of the students
were aware of what the “correct” answer should be according to the explanation, but
still five of the seven refused to accept this conclusion. Of the two who accepted the
conclusion of the explanation, one (Tl) indicated that he believed “resistance” to be
a better word than “force” to describe the effect of the table on the book, and the other
(T6) indicated that she viewed the explanation as helping “only a little” to make the
idea of an upward force from the table make sense. Thus there is strikingly little
24 BROWN

evidence that the text excerpt was helpful in improving these students’ conceptions of
the ability of static objects to exert forces.
Bridging Students. By contrast, all seven of the bridging students who initially
answered the table problem incorrectly answered the post-question about the book on
the table correctly. The following is a transcript segment in which the student gave
two reasons for the existence of a force from the table: (a) an argument from
necessity-the book presses on the table so the table musr exert a force back to relieve
the stress; and (b) an argument from mechanism-the molecules of the table are
springy, thus providing the agency (springiness) which is responsible for the force.

B2: The reason that I’m sure I’m right is that I know that the book is applying
stress upon the table and it is applying force upon the table, and hence the
table must exert the same amount of force back onto the book to relieve its
stress and readjust. And I also know that the molecules of the table are
springy and flex.

The bridging students’ responses indicated that the bridging explanation had an
effect on their conceptions about forces from static objects. Four of the students gave
an explanation indicating the agency responsible for the force (the table’s springiness),
two gave an argument from necessity indicating that a force must exist (e.g., since
the book does not fall), and one (the student just quoted) gave both types of argument
as reasons for why the table exerts an upward force.

Steel Blocks Problem

Included in the post-questions were two new questions. One of these was the steel
blocks problem about a 200-pound block resting on a @-pound block. This is a difficult
problem in that most students find it hard to imagine how a 40-pound block can exert
200 pounds of force and are thus drawn toward the conclusion that the forces between
the blocks are unequal.
The steel blocks problem is one which requires a Newtonian view of forces arising
only from an interaction between two objects. An alternative, a view which many
students apparently hold, is that force can be an innate or acquired property of an
object due, for example, to its weight or state of motion (Brown, 1989). Under the
latter view, the larger block resting on top of the smaller block would exert the greater
force due to its greater weight.
Text Excerpt Students. Although three of the seven text excerpt students said
the lower block would be exerting a force, none referred to Newton’s third law (which
indicates equal forces), even though they had just read the text excerpt about the third
law. They apparently preferred to rely on their naive conceptions of force. In the
following example segment, T1 displayed very clearly that he regarded force as an
innate or acquired property of an object; that is, that objects can “have” force.

T1: I think it [.lo-pound block B on the bottom] exerts a force up, but I don’t
think it exerts enough to stop A [200-pound block on the top] from pushing
B into the ground. See, it just makes the thing slower. So say B only weighed
one pound, then A would have 199 pounds more than B would, and so it
would push it into the ground faster. But this way, B has some force . . .
 USING EXAMPLES To REMEDIATE MISCONCEFTIONS 25

but not enough to keep A from pushing it down into the ground , . . Hard
to think about this one because in the ones before where the light thing was
on top, the heavy thing just used enough to fend off, you know, to keep the
lighter thing on top. See, so it’s a matter of how much force the thing uses.
So I’d say that, uh, A and B exert a force on each other, but A exerts a
larger force.

Bridging Srudenrs. By contrast, even though all but one of the bridging students
indicated in their explanations that they were bothered by the conception of force as
a property, six of the seven resisted the conclusion based on that assumption and
answered correctly. The following transcript segments provide examples of how the
bridging explanation can help students in this and similar problems. Although the first
student initially answered correctly, he struggled with the conception of force as B
property and was unsure whether the 40-pound block could exert 200 pounds of force.
However, his confusion was dispelled when he thought of the book resting on the
spring.

B2: Alright, I’m having trouble with this one because I’m thinking in terms of
they both should exert . . . forces on each other because B [the 40-pound
block on the bottom] has to . . . relieve that 200-pound stress. However, it
only weighs 40 pounds. Because of that number, um, I don’t know whether
it can do that. , . . Makes some sense to me. The reason [equal forces]
doesn’t make perfect sense to me is because block A is so much more heavier
than the other. Wait a minute . . . I, I’ll have to change that. I’ve just thought
about the instances of the book and the spring and of course the spring was,
weighed so much less than the book, but still the spring did bounce, the
spring did bounce back. Those atoms are still springy . . . Even if one weighs,
even if one weighs so much more than the other because sure, the book
weighed so much more than the spring, but the spring did bounce, the spring
bounded back; why can’t the same thing happen to this?

The second student below seemed to find that the “spring theory” (apparently,
that each block would “give” on contact) was helpful in thinking about the forces each
block would exert on the other.

B5: Each exerts a force, and these forces are the same size. Yes. Uh, ‘cause the
B block weighing, uh, let’s see, they would give a little, and so would the
A block; so, yeah, they’re both exerting force upon each other, going back
to the spring theory. Um, I’m sure I’m right.

In conclusion, it appears that the concept of force as an innate or acquired property

of objects is both widespread and resistant to change. All three of the text students
who answered the first part of the steel blocks problem correctly (that the 40-pound
block B does exert a force) based their answers for the second part of the question on
this assumption. Six of the seven bridging students indicated that they considered this
conception of force while answering the second part of the steel blocks problem.
However, the fact that six of the seven went on to answer both parts of the problem
correctly suggests that they had a way of thinking about the situation that enabled
them to veer away from this view of force to the Newtonian view of forces arising
from an interaction.
26 BROWN

Factors Contributing to Group Differences

In the following sections I examine protocol data indicating some possible factors
contributing to the differences in students’ responses to the two methods of using
examples. The three factors identified are (a) examples must make sense to the students,
(b) analogy relations may need to be explicitly developed between examples, and (c)
it may be necessary to develop mechanistic models which make sense to students
before they will change their conception of a situation.

Examples Must Make Sense to the Students

The scale used when students were asked to rate how much sense something made
went from one (makes no sense) to five (makes perfect sense). Students rated several
of the examples in the text excerpt as making little sense. By contrast, students indicated
that the examples in the bridging explanation made a great deal of sense. Two transcript
segments below illustrate typical student responses for the two examples of the ground
pushing forward on the runner (in the text excerpt) and the spring pressing up on the
hand (in the bridging explanation).

I: Does it make sense to you that the ground pushes forward on the athlete?
T4: No, because why should the ground all of a sudden just spontaneously
decide to push forward when there’s somebody ninning on it, but it doesn’t
push forward when there’s nobody running on it?
I: Does it make sense to you that the spring would push up on your hand?
B 4 Yeah, I can reason that out.
I: What sense rating would you give there?
B4: Five.
I: Five. Could you say a little bit about what’s behind your sense rating?
B4: Well, I know that certain metals are more, um, have elastic qualities to
them and they’re able to be bent, and they desire 10 go back to their original
position so they’ll exert however much force they’re capable of to, to try
to get back to that previous position, and then the force I’m putting on it
will move it down, and the metal structure, assuming it’s a metal spring,
will try to retain its old position.

The mean sense rating for the text excerpt examples was 3.27 out of 5, where-
as the mean sense rating for the bridging examples was significantly higher at 4.60
[t(78) = 6.19, p < 0.00011. However, as the following section shows, simply having
good individual examples, although important, may not be sufficient for understanding.

Need to Develop Analogy Relations Explicitly

Many teachers and textbook authors supplement their presentations with analogies.
However, perhaps because the situations are to them “obviously” analogous, no attempt
is made to explicitly develop the analogy relations. The present study indicates that
the use of concrete examples in this way may be ineffective. For example, even though
the physicist views the book on the table and the hand on the spring as analogous
situations, six of the seven bridging students did not. An example is given below.
 USING EXAMPLES TO REMEDIATE MISCONCEPTIONS 21

I: Is this [the hand on the spring] different from the book on the table?
B1: Yeah, because if you were pressing down on a table, couldn’t you lose
some of your force? Like just relax a little bit, your hand’s not going to
go springing back up like on, like if you were pressing down on a spring.
If you were pressing down on a spring, you press down, and you let go
just like, a little bit, like you eased up, your hand would just go up, ah,
it would be pushed up.

However, when the analogy relation between the hand on the spring and the book
on the table was developed, every student indicated that this analogy of the hand on
the spring was helpful in making the idea of an upward force from the table make
sense. The following student was asked which examples helped the idea of an upward
force from the table make sense and which did not help. He began to indicate that he
believed the spring analogy did not help, but then he realized that a way had been
“built up” from the spring to the table by means of intermediate analogies. He thus
indicated that he was aware of how the bridging strategy helped him make sense of
the table situation, even though this strategy was never described to him.

B7: I don’t think the spring, well, I guess I didn’t think the spring helped, but
in context I guess, out of context you just compare the spring and the table
it wouldn’t help, but you sort of built a way up from the spring, which is
obvious, to a flexible board, to a not so flexible board, to foam rubber, to
a table, which is pretty good.

The following points indicate that the establishment of analogical connections was
important and helpful to learning. First, for six of the seven students, the single analogy
of the hand on the spring was not sufficient to produce change, even though it made
a great deal of sense to all of the students. These six students indicated that they did
not initially view the hand on the spring as analogous to the book on the table. Second,
after some analogical connections had been developed in the explanation, all of the
students indicated that they saw at least one of the later situations, before the molecular
model, as analogous to the book on the table. This indicates that the bridging strategy,
apart from the molecular model, was helpful for these students. Third, in retrospective
comments students specifically indicated that they found the development of analogical
connections to be important. Finally, in two cases, students indicated in these retrospective
comments that the ordering of examples was important. One student even suggested
a change in the ordering to make the teaching strategy more effective.

Mechanistic Models are Important

The bridging explanation also helped students construct a mental model that
provides a mechanism indicating how the table can exert a force-the microscopic
compression or bending of the table. Such a mechanistic model was lacking in the
text excerpt. The absence of a source or agency for the force troubled several of the
students in the text group, a sentiment pointedly illustrated by the segment below:

I: Can you summarize the main idea of this explanation?

T4: Um, well they’re trying to tell me that, um, for every force there’s an
opposite force that happens against it. But they still haven’t told me where
28 BROWN

it comes from or why, and I have no intention of accepting it until they

do.

The absence of a mechanistic model may have led students to think about force in
their usual way-as an innate or acquired property of an object-rather than arising
as a result of an interaction. Some examples of such thinking in the steel blocks
problem were presented earlier.
It appears that when a force is said to be present in a given situation, students
want to know where that force is coming from, that is, what agency is responsible
for the force. Unlike the text excerpt, the bridging explanation attempted to answer
that question with the model of the table as springy and with the deeper model of the
table as composed of molecules connected by springy bonds. An example of the efficacy
of such a mechanistic model follows:

I: Let me ask if you could just rate how much does the explanation help the
idea of an upward force from the table make sense, on a scale from one
to five.
B4: Yeah, five, because there, there’s an explanation of the molecular makeup
and the reasoning behind why the table is actually . . . providing a force
back on a molecular, at a small scale, the, even the stiff board and the stiff
table is providing a force upward . . . The molecular makeup’s good far
reasoning the whole thing out.

The bridging explanation provided a source for the force from the table-the
table’s small amount of springiness or “bendiness”-thus drawing on students’ causal
reasoning and intuitions in a way that the text excerpt did not.

Discussion
Although examples are generally recognized as important in teaching conceptual
material, little research has been conducted on the use of examples in the presence of
misconceptions. A question of particular concern is whether a teaching-by-example
technique will be effective when students hold a misconception. The study reported
here indicates that the use of examples in the traditional teaching-by-example format
can be ineffe~tive.~ Although students may, to some extent, form the concept desired,
they may also remain tentative unless and until they can understand why the concept
should be applied to particular situations. Until they can make sense of their learning
in this way, some students may refuse to accept that a particular principle or concept
applies in certain situations. These results are consistent with recent work on conceptual
cohesiveness indicating that an individual’s sense of which examples should be grouped
together under a given concept or principle is governed largely by a person’s naive
theories and models (Medin & Ortony, 1989; Medin & Wattenmaker, 1987; Murphy
& Medin, 1985).

’ It is worth noting that the text excerpt contains no nonexamples of Newton’s third law. Although
this is undesirable from the perspective of an optimal inductive use of examples, typically one would expect
overgeneralization from using only examples of a principle, not the defiant undergeneralization observed
here.
 USING EXAMPLES TO REMEDIATE MISCONCEF‘TIONS 29

The students who received the text excerpt apparently did not treat the examples
as neutral representations from which they could abstract common features. Rather,
they seemed to reason causally about the situations using their existing conception of
force, deciding on the basis of their intuitive beliefs whether forces would be present
in the examples. The bridging explanation was more effective perhaps because it
engaged and channeled the students’ causal reasoning, showing how an aspect of their
conception of force could be applicable to the table situation-a table with the agency
of springiness enabling it to exert a force, much like a prototypical example of a hand
pressing down on a spring.
However, the text excerpt did not similarly attempt to channel the student’s causal
reasoning. Rather, it treated the concrete situations as examples of an abstract principle
demonstrating the types of situations to which that principle applies. There are indications
that, for many students, there was not a successful process of induction for generating
or confirming an abstract schema in a form that could be applied to transfer problems.
Evidence from the current study indicates three factors which may contribute to
the observed differences in student response to the two explanations.
1. Some of the individual examples in the text excerpt were counterintuitive to
many students (e.g., the runner and the stone). However, all examples in the bridging
explanation appeared to be understandable to the students. In particular, all students
said that the anchoring example of the spring pushing on the hand made sense to them
intuitively.
2. In some cases, examples in the text excerpt made sense to the students by
tapping their intuition (e.g., the rifle kick), but students could not see a connection to
the book-on-the-table situation. However, the bridging explanation placed emphasis
on developing such connections by presenting the analogous cases as a connected
sequence of examples.
3. In the text excerpt, students were left wondering how the table could possibly
exert a force. Helping the student construct a mechanistic model of a situation evoking
a misconception can be an important step in helping a student change his or her
conception of the situation. Some students may even require such a model which makes
sense to them before they will change their conception of a ~ituation.~
This study raises some questions for further investigation. First, is the failure of
an inductive use of examples, as observed here, to be expected in other domains?
Second, is such a failure the norm or is it to be expected only in rare cases? Third,
in what situations is model construction via multiple analogies a viable alternative?
Although the present study was confined to a narrow domain, there is reason to believe
such a failure of the traditional approach might occur in a number of other domains
in which deep-seated preconceptions exist. In such instances, existing conceptions
could interfere with a process of schema induction via multiple examples in a traditional

It is worth noting that in a fairly recent major work on induction (Holland, Holyoak, Nisbett, &
T h a g d , 1986), the authors define induction as . . all inferential processes that expand knowledge in
‘I.

the face of uncertainty” (p. 1). Further, they state that “. . . mental models are the major source of inductive
change in long-term knowledge structures . . . the construction of mental models provides the force and
direction for the process of induction” (p. 14). However, their definitionsof “induction” and “mental model”
are much broader than the conceptions of “schema induction” and “mechanistic conceptual model.” See
Brown and Clement (1989) and Clement (1989) for more extended treatments of induction and model
construction as employed here.
30 BROWN

teaching-by-example approach. When this is the case, an approach which draws on

students’ existing valid intuitions and helps them extend these intuitions analogically
in a process of model construction may be a viable and effective alternative to the
traditional approach. Further research is needed to provide more conclusive answers
to these questions.

Conclusions
The present study indicates that the use of concrete examples can be an effective
means for inducing conceptual change. However, if the conclusions of this study are
confirmed in other domains, this means that the particular method one uses in example-
based teaching can be crucial to learning outcomes. Teachers need to be aware that
examples which they find compelling may not be at all illuminating for their students.
Even when an example is compelling to a student, it may not be seen as analogous
to target problems drawing out a misconception. Such analogicalconnections of qualitative
similarity are not always obvious, and may require attention in instruction through
techniques such as bridging. Also, teachers need to focus on the goal of helping
students develop visualizable, qualitative, mechanistic models of physical phenomena
which can help students to make sense of the more abstract principles often invoked
to explain the phenomena.
Apparently, when students hold a misconception, not all examples (or ways of
using examples) are equally effective. When the examples make sense to the student,
are related analogically to target problems, and help the student form a qualitative
mental model, the examples may be more helpful than when they simply show the
range of application of an abstract concept or principle.

Author Notes

This article contains material drawn from my doctoral dissertation, Using Analogies and
Examples to Help Students Overcome Misconceptions in Physics: A Comparison of Two Teach-
ing Strategies, Brown (1987). Preparation of this article was supported by National Science
Foundation Grants No. MDR-8470579 and No. MDR-8751398. I would particularly like to
thank John Clement and Klaus Schultz for their insightful contributions to the ideas presented
in this article. I would also like to thank Arnold Well, Pamela Hardiman, and Clifford Konold
for their constructive comments on earlier drafts of this article. For the past five years I have
been involved with a project investigating students’ conceptual understanding in physics and
ways of remediating misconceptions. Requests for reprints should be sent to David Brown,
Department of Cumculum and Instruction, 311 Education Building, 1310 South Sixth St.,
Champaign, IL 61820-6990.

Appendix A

Text Excerpt
In this exercise we will consider the question of whether a table pushes up on a
book resting on it. Newton’s third law says that the table does exert a force on the
book. Newton’s third law states: To every action there is always opposed an equal
 USING EXAMPLES To REMEDIATE MISCONCEPTIONS 31

reaction: or, mutual actions of two bodies upon each other are always equal and directed
to contrary parts. This is a word-for-word translation from the Principia. In modern
usage, however, we would useforce where Newton used the Latin word for action.
So we could rewrite this passage as follows: If one object exerts a force on another,
then the second also exerts a force on the first; these forces are equal in magnitude
and opposite in direction.
Apply this idea to an athlete running. You now see that her act of pushing with
her feet back against the ground (call it the action) also involves a push of the ground
forward on her (call it the reaction). It is this reaction that propels her forward.
In this and all other cases, it really makes no difference which force you call the
action and which the reaction, because they occur at exactly the same time. The action
does not “cause” the reaction. If the earth could not “push back” on her feet, the
athlete could not push on the earth in the first place. Instead, she would slide around
as on slippery ice. Action and reaction coexist. You cannot have one without the other.
Most important, the two forces are not acting on the same body. In a way, they are
like debt and credit. One is impossible without the other; they are equally large but
of opposite sign, and they happen to two different objects.
Newton wrote: “Whatever draws or presses another is as much drawn or pressed
by that other. If you press a stone with your finger, the finger is also pressed by the
stone.” This statement suggests that forces always arise as a result of mutual actions
(“interactions”) between objects. If object A pushes or pulls on B, then at the same
time object B pushes or pulls with precisely equal force on A. These paired pulls and
pushes are always equal in magnitude, opposite in direction, and on two different
objects.
Every day you see hundreds of examples of this law at work. A boat is propelled
by the water that pushes forward on the oar while the oar pushes back on the water.
A car is set in motion by the push of the ground on the tires as they push back on the
ground; when friction is not sufficient, the push on the tires cannot start the car forward.
While accelerating a bullet forward, a rifle experiences recoil, or “kick.” A balloon
shoots forward while the air spurts out from it in the opposite direction. Many such
effects are not easily observed. For example, when an apple falls, pulled down by its
attraction to the earth, i.e., by its weight, the earth, in turn, accelerates upward slightly,
pulled up by the attraction of the earth to the apple.
To summarize, many people say the table is not exerting a force upward on the
book. However, the book is exerting a force downward on the table because of its
weight. Therefore, because of Newton’s third law, the table is exerting an equal force
upward on the book.

Bridging Explanation
In this exercise we will consider the question of whether a table pushes up on a
book resting on it. Consider pushing down on a spring with your hand.
Now consider the case of a heavy dictionary being placed on a bedspring so the
spring compresses some.
When the book is placed on the spring, the spring compresses. The further down
the spring is pushed, the more it pushes back. The spring is compressed by the book
to the point where it pushes back with a force equal to the book’s weight. For example,
if the book weighs 10 pounds, the spring compresses until it exerts an equal upward
32 BROWN

force of 10 pounds. In a similar way, if you hold a :30-pound dictionary in your

outstretched hand, you have to exert an upward force of 30 pounds to hold it there.
Many people say the book on the spring is different than the book on the table.
They say that although neither is alive, the spring compresses but the table is rigid.
But is the table rigid? Imagine a flexible board between two sawhorses. If you were
to push down on this board it would bend and push back, just like pushing down on
the spring. The board would also push back on a book, just like the spring. Now
imagine thicker and thicker boards.
If you had a thick enough board, it would be just like a table. Both the board and
the table would bend a tiny, tiny bit under the weight of a book. Another way to think
of the table is like very stiff foam rubber. Even though the stiff foam rubber would
not compress much under the weight of a book, it would compress some.
The table is composed of molecules which are connected to other molecules by
bonds which are “springy.” Thus the table has some amount of give or “bendiness”
or “squishiness” to it. If you were to look closely with a microscope you would see
that the book causes a slight depression in the table. The table, just like the spring,
the flexible board, or foam rubber, is bent or compressed some and thus pushes back.
Like the spring holding the dictionary, the table bends or compresses just enough to
provide an upward force equal to the book’s weight.
To summarize, many people do not think the table can exert a force since it is
rigid and lifeless. However, they feel a spring can exert a force if a force is exerted
on it because it “wants to get back to its original shape.” Thus there seems to be a
distinction between rigid objects and springy objects. However, if you look closely
enough at a table it is springy because of its molecular makeup. Because of this springy
nature of all matter, the table can and does exert a force upward on the book. Just
like a spring, the table compresses (on a miscroscopic scale) until it is compressed
enough to provide an upward force equal to the book’s weight.

Appendix B

Questions
Table Problem (Pre- and Post-). A book is at rest on a table. Does the table
exert an upward force on the book?
Goat Problem (Pre- and Post-). A stubborn goat is pushing against a wall. While
the goat is pushing, does the wall exert a force back on the goat?
Mosquito Problem (Pre- and Post-). On a day with no wind, a mosquito lands
on top of the Washington Monument. Think about whether the mosquito exerts force
on the monument and whether the monument exerts a force on the mosquito while it
is resting there. While the mosquito is resting there, does the monument exert an
upward force on the mosquito’?
Two Boxes Problem (Post- Only). A box weighing 50 pounds rests on top of
another box weighing 100 pounds. Think about whether the upper box exerts a force
on the lower box and whether the ground exerts a force on the lower box. Does the
ground exert an upward force on the lower box?
 USING EXAMPLES TO REMEDIATE MISCONCEPTIONS 33

Steel Blocks Problem (Post- Only). A large steel block (A) weighing 200 pounds
rests on a small steel block (B) weighing 40 pounds. Think about whether A exerts a
force on B and whether B exerts a force on A. Does B exert an upward force on A?

References
Brown, D. (1987). Using analogies and examples to help students overcome
misconceptions in physics: A comparison of two teaching strategies. Dissertation
Abstracts International, 49, 473A. (University of Microfilms No. 8805897).
Brown, D. (1989). Students’ concept of force: The importance of understanding
Newton’s third law. Physics Education, 24, 353-358.
Brown, D., & Clement, J. (1987). Misconceptions concerning Newton’s law of
action and reaction: The underestimated importance of the third law. In J. D. Novak
(Ed.), Proceedings of the Second International Seminar, Misconceptions and Educational
Strategies in Science and Mathematics (Vol. 3 , pp. 39-53).
Brown, D., & Clement, J. (1989). Overcoming misconceptions via analogical
reasoning: Abstract transfer versus explanatory model construction. lnsmuctional Science,
18, 237-261.
Champagne, A.B., Gunstone, R.F., & Klopfer, L.E. (1982, May). A perspective
on the differences between expert and novice performance in solving physics problems.
Paper presented at a meeting of the Australian Science-EducationResearch Association,
Sydney, Australia.
Champagne, A.B., Klopfer, L.E., & Anderson, J.H. (1980). Factors influencing
the learning of classical mechanics. American Journal of Physics, 50, 1074- 1079.
Clement, J. (1982). Students’ preconceptions in introductorymechanics. American
Journal of Physics, 50, 66-7 1.
Clement, J. (1989). Learning via model construction and criticism: Protocol evidence
on sources of creativity in science. In J. Glover, R. Ronning, & C. Reynolds (Eds.),
Handbook of creativity: Assessment, theory, and research (pp. 341-381). New York:
Plenum.
Driver, R., & Easley, J. (1978). Pupils and paradigms: A review of literature
related to concept development in adolescent science students. Studies in Science
Education, 5 , 61-84.
Driver, R., & Erickson, G . (1983). Theories-in-action: Some theoretical and
empirical issues in the study of students’ conceptual frameworks in science. Studies
in Science Education, 10, 37-60.
Gentner, D. (1983). Structure mapping: A theoretical framework for analogy.
Cognitive Science, 7 , 155- 170.
Gentner, D. (1989). The mechanisms of analogical learning. In S. Vosniadou &
A. Ortony (Eds.), Similarity and analogical reasoning (pp. 199-241). New York:
Cambridge University Press.
Gick, M.L., & Holyoak, K.J. (1983). Schema induction and analogical transfer.
Cognitive Psychology, 15, 1-38.
Halloun, L A . , & Hestenes, D. (1985). The initial knowledge state of college
physics students. American Journal of Physics, 53, 1043- 1055.
Holland, J.H., Holyoak, K.J., Nisbett, R.E., & Thagard, P.R. (1986). Induction:
Processes of inference, learning, and discovery. Cambridge, MA: MIT Press.
34 BROWN

Kaiser, M.K., Jonides, J., & Alexander, J. (1986). Intuitive reasoning about
abstract and familiar physics problems. Memory & Cognition, 14, 308-312.
McCloskey, M. (1983). Intuitive physics. Scientific American, 248, 122- 130.
McDermott, L. (1984, July). Research on conceptual understanding in mechanics.
Physics Today, 37, 24-32.
Medin, D., & Ortony, A. (1989). Psychological essentialism. In S. Vosniadou
& A. Ortony (Eds.), Similarity and analogical reasoning (pp. 179-195). New York:
Cambridge University Press.
Medin, D.,& Wattenmaker, W.D. (1987). Category cohesiveness, theories, and
cognitive archeology. In U. Neisser (Ed.), Concepts and conceptual development:
Ecological and intellectual factors in categorization (pp. 25-62). New York: Cambridge
University Press.
Mervis, C.B., & Pani, J.R. (1980). Acquisition of basic object categories. Cognitive
Psychology, 12, 496-522.
Murphy, G.L., & Medin, D.L. (1985). The role of theories in conceptual coherence.
Psychological Review, 92, 289-3 16.
Rosch, E. (1973). On the internal structure of perceptual and semantic categories.
In T. Moore (Ed.), Cognitive development and the acquisition of language. New York:
Academic Press.
Rosch, E., & Mervis, C. (1975). Family resemblances: Studies in the internal
structure of categories. Cognitive Psychology, 7 , 573-605.
Rutherford, F.J., Holton, G., & Watson, F.G. (Eds.) (1981). Project physics.
New York: Holt, Rinehart, & Winston.
Sjoberg, S., & Lie, S. (198 1). Ideas about force and movement among Norwegian
pupils and students (Tech. Rep. No. 81-11). Oslo, Norway: University of Oslo.
Tennyson, R.D., & Park, 0. (1980). The teaching of concepts: A review of
instructional design research literature. Review of Educational Research, 50, 55-70.
Viennot, L. ( 1979). Spontaneous reasoning in elementary dynamics. European
Journal of Science Education, 1, 205-22 1.

Manuscript accepted April 6, 1990.